Bug 1143818 - [QE] (6.4.0) Shutting down server with deployed JMS bridge aborts XA transactions
Summary: [QE] (6.4.0) Shutting down server with deployed JMS bridge aborts XA transact...
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: JBoss Enterprise Application Platform 6
Classification: JBoss
Component: HornetQ
Version: 6.4.0
Hardware: Unspecified
OS: Unspecified
unspecified
medium
Target Milestone: DR9
: EAP 6.4.0
Assignee: Clebert Suconic
QA Contact: Miroslav Novak
Russell Dickenson
URL:
Whiteboard:
Depends On:
Blocks: 1158079
TreeView+ depends on / blocked
 
Reported: 2014-09-18 05:22 UTC by Miroslav Novak
Modified: 2014-12-18 21:49 UTC (History)
5 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2014-12-15 16:35:18 UTC
Type: Bug
Embargoed:

Attachments (Terms of Use)
server.log (225.06 KB, text/plain)
2014-09-18 05:23 UTC, Miroslav Novak 		no flags Details
standalone-full-ha-bridge.xml (28.91 KB, text/plain)
2014-09-18 05:24 UTC, Miroslav Novak 		no flags Details
standalone-full-ha-target.xml (27.70 KB, text/plain)
2014-09-18 05:24 UTC, Miroslav Novak 		no flags Details
standalone-full-ha-2.xml (28.09 KB, application/xml)
2014-09-22 08:04 UTC, Miroslav Novak 		no flags Details
Show Obsolete (1) View All


Links
System ID Private Priority Status Summary Last Updated
Red Hat Bugzilla 1174338 0 unspecified CLOSED [Doc Bug Fix] Add note - if JMS bridge is configured with quality-of-server ONCE_AND_ONLY_ONCE then shutdown server with... 2021-02-22 00:41:40 UTC
Red Hat Issue Tracker HORNETQ-1440 0 Major Closed BZ-1143818 - reset connected flags after failover 2014-10-30 16:07:16 UTC

Internal Links: 1174338

Description Miroslav Novak 2014-09-18 05:22:19 UTC 
Description of problem:
If server with deployed JMS bridge (ONCE_AND_ONLY_ONCE delivery mode) is processing messages and this server is cleanly shutdowned than there are lots of arjuna warnings (see server.log)

Steps to Reproduce:
1. Start two EAP 6.4.0.DR1 servers and deploy JMS bridge to first. Bridge sends messages from InQueue (from 1st server) to OutQueue (2nd server)
2. Start producer which sends messages to InQueue
3. Cleanly shutdown first server - see warnings

Expected results:
Before shutdown the bridge should stop to process new messages from InQueue and wait some time for current messages to be processed.
Comment 1 Miroslav Novak 2014-09-18 05:23:27 UTC 
Created attachment 938745 [details]
server.log
Comment 2 Miroslav Novak 2014-09-18 05:24:11 UTC 
Created attachment 938746 [details]
standalone-full-ha-bridge.xml
Comment 3 Miroslav Novak 2014-09-18 05:24:51 UTC 
Created attachment 938747 [details]
standalone-full-ha-target.xml
Comment 4 Miroslav Novak 2014-09-18 08:02:57 UTC 
After some tries I've noticed that steps to reproduce the issue are different:
1. Start two EAP 6.4.0.DR1 servers and deploy JMS bridge to first. Bridge sends messages from InQueue (from 1st server) to OutQueue (2nd server)
2. Cleanly shutdown 2nd server and start again
3. Cleanly shutdown first server - see warnings
Comment 5 Miroslav Novak 2014-09-18 08:05:02 UTC 
@Andy
I'd like to ask you to take a look at this first before it's moved to another component.
Comment 6 Andy Taylor 2014-09-18 10:56:11 UTC 
yes, this is for us so leave it with me
Comment 7 Miroslav Novak 2014-09-22 08:04:06 UTC 
Created attachment 939911 [details]
standalone-full-ha-2.xml

New standalone-full-ha-2.xml for target server of the bridge attached.
Comment 8 Andy Taylor 2014-09-22 14:07:56 UTC 
PR created https://github.com/hornetq/hornetq/pull/1858.

Just to note that if the target server is down then an exception will still be shown. This is only in XA mode as we batch messages and need to rollback the transaction. We should make sure that this is documented and that in this scenario the target server should be stopped after the source server (or other way around if the source server is remote).
Comment 10 JBoss JIRA Server 2014-10-30 16:07:16 UTC 
Clebert Suconic <clebert.suconic> updated the status of jira HORNETQ-1440 to Closed
Comment 11 Ivo Studensky 2014-12-11 11:44:42 UTC 
This has been fixed in 2.3.22.Final, see comment 9.
Comment 13 Miroslav Novak 2014-12-15 08:39:49 UTC 
I can see exceptions in EAP 6.4.0.DR13 (HQ 2.3.24.Final). Returning to assigned.

Based on comment #8 this should be closed as won't fix, right?
Comment 14 Andy Taylor 2014-12-15 16:18:52 UTC 
if in xa mode yes
Comment 15 Miroslav Novak 2014-12-15 16:35:18 UTC 
Ok, closing as won't fix and creating new bz for doc team bz#1174338.
Note You need to log in before you can comment on or make changes to this bug.